536 Commits

Author SHA1 Message Date
Jedrzej Kosinski
6ea69369ce Merge branch 'master' into worksplit-multigpu 2025-08-07 23:24:02 -07:00
comfyanonymous
9126c0cfe4
Qwen Image model merging node. (#9202) 2025-08-06 04:07:04 -04:00
Jedrzej Kosinski
df122a7dba Merge branch 'master' into worksplit-multigpu 2025-08-01 12:31:57 -07:00
comfyanonymous
5ee381c058
Fix WanFirstLastFrameToVideo node when no clip vision. (#9134) 2025-07-31 23:33:27 -04:00
Jedrzej Kosinski
67e906aa64 Merge branch 'master' into worksplit-multigpu 2025-07-31 04:00:22 -07:00
guill
0a3d062e06
ComfyAPI Core v0.0.2 (#8962)
* ComfyAPI Core v0.0.2

* Respond to PR feedback

* Fix Python 3.9 errors

* Fix missing backward compatibility proxy

* Reorganize types a bit

The input types, input impls, and utility types are now all available in
the versioned API. See the change in `comfy_extras/nodes_video.py` for
an example of their usage.

* Remove the need for `--generate-api-stubs`

* Fix generated stubs differing by Python version

* Fix ruff formatting issues
2025-07-29 22:17:22 -04:00
Jedrzej Kosinski
5d5024296d Merge branch 'master' into worksplit-multigpu 2025-07-28 06:17:24 -07:00
comfyanonymous
a88788dce6
Wan 2.2 support. (#9080) 2025-07-28 08:00:23 -04:00
Jedrzej Kosinski
3b90a30178 Merge branch 'master' into worksplit-multigpu-wip 2025-07-27 01:03:25 -07:00
Eugene Fairley
4293e4da21
Add WAN ATI support (#8874)
* Add WAN ATI support

* Fixes

* Fix length

* Remove extra functions

* Fix

* Fix

* Ruff fix

* Remove torch.no_grad

* Add batch trajectory logic

* Scale inputs before and after motion patch

* Batch image/trajectory

* Ruff fix

* Clean up
2025-07-24 20:59:19 -04:00
Kohaku-Blueleaf
eb2f78b4e0
[Training Node] algo support, grad acc, optional grad ckpt (#9015)
* Add factorization utils for lokr

* Add lokr train impl

* Add loha train impl

* Add adapter map for algo selection

* Add optional grad ckpt and algo selection

* Update __init__.py

* correct key name for loha

* Use custom fwd/bwd func and better init for loha

* Support gradient accumulation

* Fix bugs of loha

* use more stable init

* Add OFT training

* linting
2025-07-23 20:57:27 -04:00
Jedrzej Kosinski
3c4104652b Merge branch 'master' into worksplit-multigpu-wip 2025-07-22 11:42:23 -07:00
comfyanonymous
0aa1c58b04
This is not needed. (#8991) 2025-07-21 16:48:25 -04:00
comfyanonymous
54a45b9967
Replace torchaudio.load with pyav. (#8989) 2025-07-21 14:19:14 -04:00
chaObserv
100c2478ea
Add SamplingPercentToSigma node (#8963)
It's helpful to adjust start_percent or end_percent based on the corresponding sigma.
2025-07-19 23:09:11 -04:00
comfyanonymous
1b96fae1d4
Add nested style of dual cfg to DualCFGGuider node. (#8965) 2025-07-19 04:55:23 -04:00
comfyanonymous
7f492522b6
Forgot this (#8957) 2025-07-18 05:43:02 -04:00
comfyanonymous
650838fd6f
Experimental CFGNorm node. (#8942)
This is from the new hidream e1 1 model code. Figured it might be useful as a generic cfg trick.
2025-07-17 04:11:07 -04:00
Brandon Wallace
6b8062f414
Fix MaskComposite error when destination has 2 dimensions (#8915)
Fix code that is using the original `destination` input instead of the reshaped value.
2025-07-15 21:08:27 -04:00
comfyanonymous
9ca581c941
Remove windows line endings. (#8902) 2025-07-14 13:10:20 -04:00
chaObserv
b43916a134
Fix fresca's input and output (#8871) 2025-07-11 12:52:58 -04:00
Kohaku-Blueleaf
1205afc708
Better training loop implementation (#8820) 2025-07-09 11:41:22 -04:00
kosinkadink1@gmail.com
9855baaab3 Merge branch 'master' into worksplit-multigpu 2025-07-09 03:57:30 -05:00
Kohaku-Blueleaf
181a9bf26d
Support Multi Image-Caption dataset in lora training node (#8819)
* initial impl of multi img/text dataset

* Update nodes_train.py

* Support Kohya-ss structure
2025-07-08 20:18:04 -04:00
chaObserv
aac10ad23a
Add SA-Solver sampler (#8834) 2025-07-08 16:17:06 -04:00
comfyanonymous
e740dfd806
Fix warning in audio save nodes. (#8818) 2025-07-07 03:16:00 -04:00
comfyanonymous
7eab7d2944
Remove dependency on deprecated torchaudio.save function (#8815) 2025-07-06 14:01:32 -04:00
comfyanonymous
f74fc4d927
Add ImageRotate and ImageFlip nodes. (#8789) 2025-07-03 19:16:30 -04:00
City
d9277301d2
Initial code for new SLG node (#8759) 2025-07-02 20:13:43 -04:00
comfyanonymous
34c8eeec06
Fix ImageColorToMask not returning right mask values. (#8771) 2025-07-02 15:35:11 -04:00
Harel Cain
9f1069290c
nodes_lt: fixes to latent conditioning at index > 0 (#8769) 2025-07-02 15:34:51 -04:00
Terry Jia
79ed752748
support upload 3d model to custom subfolder (#8597) 2025-07-01 20:43:48 -04:00
Jedrzej Kosinski
d53479a197 Merge branch 'master' into worksplit-multigpu 2025-07-01 17:33:05 -05:00
comfyanonymous
772de7c006
PerpNeg Guider optimizations. (#8753) 2025-07-01 03:09:07 -04:00
chaObserv
b22e97dcfa
Migrate ER-SDE from VE to VP algorithm and add its sampler node (#8744)
Apply alpha scaling in the algorithm for reverse-time SDE and add custom ER-SDE sampler node for other solver types (SDE, ODE).
2025-07-01 02:38:52 -04:00
chaObserv
f02de13316
Add TCFG node (#8730) 2025-07-01 02:33:07 -04:00
comfyanonymous
cf49a2c5b5
Dual cfg node optimizations when cfg is 1.0 (#8747) 2025-06-30 14:18:25 -04:00
chaObserv
5b4eb021cb
Perpneg guider with updated pre and post-cfg (#8698) 2025-06-28 18:13:13 -04:00
comfyanonymous
ef5266b1c1
Support Flux Kontext Dev model. (#8679) 2025-06-26 11:28:41 -04:00
comfyanonymous
ec70ed6aea
Omnigen2 model implementation. (#8669) 2025-06-25 19:35:57 -04:00
comfyanonymous
bd9f166c12
Cosmos predict2 model merging nodes. (#8647) 2025-06-24 05:17:16 -04:00
Jedrzej Kosinski
443a795850 Merge branch 'master' into worksplit-multigpu 2025-06-24 00:49:24 -05:00
comfyanonymous
ae0e7c4dff
Resize and pad image node. (#8636) 2025-06-22 17:59:31 -04:00
comfyanonymous
78f79266a9
Allow padding in ImageStitch node to be white. (#8631) 2025-06-22 00:19:41 -04:00
Jedrzej Kosinski
1ae98932f1 Merge branch 'master' into worksplit-multigpu 2025-06-17 04:58:56 -05:00
comfyanonymous
ae75a084df
SaveLora now saves in the same filename format as all the other nodes. (#8538) 2025-06-15 03:44:59 -04:00
comfyanonymous
d6a2137fc3
Support Cosmos predict2 image to video models. (#8535)
Use the CosmosPredict2ImageToVideoLatent node.
2025-06-14 21:37:07 -04:00
Kohaku-Blueleaf
520eb77b72
LoRA Trainer: LoRA training node in weight adapter scheme (#8446) 2025-06-13 19:25:59 -04:00
comfyanonymous
5bf69bde35
Add cosmos_rflow option to ModelSamplingContinuousEDM node. (#8523)
This is for the cosmos predict2 model.
2025-06-13 17:47:52 -04:00
Kent Mewhort
c7b25784b1
Fix WebcamCapture IS_CHANGED signature (#8413) 2025-06-09 13:05:54 -04:00